Prepare MODIS Data

This code imports, cleans, and merges MODIS product data exported using the AppEEARS web application.

Exported MODIS data for the period 2000-2016:

• MOD11A2: LST Day 1km & LST Night 1km (8 day)
• MOD13Q1: EVI 250m & NDVI 250m (16 day)
• MOD15A2H: fPAR 500m & LAI 500m (8 day)
• MOD16A2: LE 500m & PET 500m (8 day)
• MOD17A2H: GPP 500m (8 day)

Data was exported for 50 BAMS paper sites (.csv format, “ID”, “Category”, “LAT”, “LONG”)

9 MODIS products are downloaded in 5 .csv files:

[1] "./BAMS-Sites-MOD11A2-006-results.csv"  "./BAMS-Sites-MOD13Q1-006-results.csv" 
[3] "./BAMS-Sites-MOD15A2H-006-results.csv" "./BAMS-Sites-MOD16A2-006-results.csv" 
[5] "./BAMS-Sites-MOD17A2H-006-results.csv"
[[1]]

[[2]]

[[3]]

[[4]]

[[5]]

NA

Simplify data by removing unnecssary fields:

$lst_day

$vi

$lai

$et

$gpp

$lst_night

NA

Look at what the data quality levels are, convert good to 1, keep only 1s

[[1]]

[[2]]

[[3]]

[[4]]

[[5]]

[[6]]

Make a plot of daytime LST to make sure things look OK

ggplot(modis.filtered[[1]], aes(DATE, LST_D, col = as.factor(LAT))) +
  geom_point()

Combine all the MODIS products for all sites and years into one dataframe

lst_d.filtered <- modis.filtered$lst_day %>% mutate(PRODUCT = rep("LST_D",length(LAT)), VAL = LST_D) %>% select(LAT, LONG, DATE, PRODUCT, VAL)
lst_n.filtered <- modis.filtered$lst_night %>% mutate(PRODUCT = rep("LST_N",length(LAT)), VAL = LST_N) %>% select(LAT, LONG, DATE, PRODUCT, VAL)
evi.filtered <- modis.filtered$vi %>% mutate(PRODUCT = rep("EVI",length(LAT)), VAL = EVI) %>% select(LAT, LONG, DATE, PRODUCT, VAL)
ndvi.filtered <- modis.filtered$vi  %>% mutate(PRODUCT = rep("NDVI",length(LAT)), VAL = NDVI) %>% select(LAT, LONG, DATE, PRODUCT, VAL)
fpar.filtered <- modis.filtered$lai  %>% mutate(PRODUCT = rep("FPAR",length(LAT)), VAL = FPAR) %>% select(LAT, LONG, DATE, PRODUCT, VAL)
lai.filtered <- modis.filtered$lai  %>% mutate(PRODUCT = rep("LAI",length(LAT)), VAL = LAI) %>% select(LAT, LONG, DATE, PRODUCT, VAL)
gpp.filtered <- modis.filtered$gpp %>% mutate(PRODUCT = rep("GPP",length(LAT)), VAL = GPP) %>% select(LAT, LONG, DATE, PRODUCT, VAL) 
le.filtered <- modis.filtered$et %>% mutate(PRODUCT = rep("LE",length(LAT)), VAL = LE) %>%  select(LAT, LONG, DATE, PRODUCT, VAL) 
pet.filtered <- modis.filtered$et %>% mutate(PRODUCT = rep("PET",length(LAT)), VAL = PET) %>% select(LAT, LONG, DATE, PRODUCT, VAL) 
modis.filtered <- rbind(lst_d.filtered,lst_n.filtered,evi.filtered,ndvi.filtered,fpar.filtered,lai.filtered,
                        gpp.filtered,le.filtered,pet.filtered)
modis.filtered$PRODUCT <- as.factor(modis.filtered$PRODUCT)
str(modis.filtered)

Classes ‘tbl_df’, ‘tbl’ and 'data.frame':   200018 obs. of  5 variables:
 $ LAT    : num  27.2 27.2 27.2 27.2 27.2 ...
 $ LONG   : num  -81.2 -81.2 -81.2 -81.2 -81.2 ...
 $ DATE   : Date, format: "2000-03-13" "2000-03-21" "2000-03-29" ...
 $ PRODUCT: Factor w/ 9 levels "EVI","FPAR","GPP",..: 6 6 6 6 6 6 6 6 6 6 ...
 $ VAL    : num  305 307 307 307 308 ...

Finally , join the observations to the Site List on the latitude column:

modis.full <- full_join(bams.sites, modis.filtered, by =c("LAT","LONG"))
str(modis.full)

'data.frame':   206698 obs. of  7 variables:
 $ ID      : Factor w/ 50 levels "BCBog","BCFEN",..: 21 21 21 21 21 21 21 21 21 21 ...
 $ Category: Factor w/ 5 levels "CRO_OTHER","CRO_RICE",..: 5 5 5 5 5 5 5 5 5 5 ...
 $ LAT     : num  68.3 68.3 68.3 68.3 68.3 ...
 $ LONG    : num  19.1 19.1 19.1 19.1 19.1 ...
 $ DATE    : Date, format: "2000-04-06" "2000-05-16" "2000-07-03" ...
 $ PRODUCT : Factor w/ 9 levels "EVI","FPAR","GPP",..: 6 6 6 6 6 6 6 6 6 6 ...
 $ VAL     : num  269 278 284 289 291 ...

Finished! Save as .csv

# set wd
setwd("C:/Users/Gavin McNicol/Box Sync/MODIS Data/AppEEARS Data/bams-sites")

The working directory was changed to C:/Users/Gavin McNicol/Box Sync/MODIS Data/AppEEARS Data/bams-sites inside a notebook chunk. The working directory will be reset when the chunk is finished running. Use the knitr root.dir option in the setup chunk to change the working directory for notebook chunks.

write.csv(modis.full, "bams_sites_MODIS_full.csv")